Proposer's Name: Mr. Parminder Jeet Singh
Proposer's Organization: Just Net Coalition
Co-Proposer's Name: Mr. Roberto Bissio
Co-Proposer's Organization: Social Watch
Co-Organizers:
Mr. Parminder Jeet, SINGH, civil society, Just Net Coalition,
Mr. Roberto, BISSIO, civil society, Social Watch,
Mr. Hans KLIEN, academic community, Georgia Tech
Valeria BETANCOURT, civil society, Association for Progressive Communications
Mishi CHOUDHARY, Technical community, Software Freedom Law Centre of India
Session Format: Round Table - 90 Min
Proposer:
Country: India
Stakeholder Group: Civil Society
Co-Proposer:
Country: Uruguay
Stakeholder Group: Civil Society
Speaker: Pretam Maloor
Speaker: Mishi Choudhary
Speaker: Hans Klein
Content of the Session:
The world had not yet gotten over its shock and awe at the power of open information flows, networking and then big data, that just within the last year or two we see that it will be artificial intelligence (AI) that could be the real game changer. AI hits at the very model of knowledge making that human civilisation has developed over millennia, which is at the base of all its evolution and development. Knowledge building was based on studying empirical facts, developing hypothesis, testing them and building theories and models of general action and prediction. Continuous micro digital mapping of human and allied activities puts so much data or 'facts' in the hands of machines that they can correlate them and find patterns that defy the most sophisticated existing model building practices. Correlation trumps causation and provides much greater predictive value, which can thus be used to control social and natural phenomenon. Whole systems of society can work pretty autonomously based on AI, which for the first time represents a wholesale disembodiment of intelligence as happened with mechanical power with the advent of the industrial revolution.
Those who control AI would be able to exert control across whole sectors, and whole of society, in ways that are unprecedented. Almost all of it is currently owned by corporation, and thus as the trends stand, AI powered society may represent a new level of corporatist re-organisation of society. A society requires both economic efficiency to maximise production as it requires political processes enforcing equity and social justice, for a just distribution of its productive outputs. AI may well solve the issue of production forever, which makes us need to focus on the processes for equity and justice. However, with near complete control over AI by a few corporates, and little political and regulatory advances in this area, it is not clear how AI will helps us more to a more equal rather than unequal society. With AI, where even the machine cannot spell out the basis of its actions other than justifying it with efficient results, the issues of ethics, equity and justice need to addressed anew, starting from conceptual levels, and building political processes and regulatory practices upon them.
This workshop will address these fundamental issues. How can human beings keep track of what AI systems are up to, what is the basis of their actions, which is necessary to anticipate and “control” them? Can some kind of ethical and regulatory super-instructions be built into all AI systems, as a politically enforced requirement, which overwrite all AI actions, however efficient they may otherwise it, and even “controls” its learning? How can these social and political imperatives override straightforward efficiency (and corporate interest) driven AI systems?
These are of course complex issues and questions that stand at the intersection of the socio-political realm and technology developments, which however, in our view, must be begun to be addressed right away as we stand at the cusp of a new technology wave, that could redefine social organisation.
Relevance of the Session:
Almost all big digital corporations have declared that AI will be core to their strategies. We are seeing corporations begin to dominate different social sectors, like transportation, e-commerce etc, increasingly employing AI. Governments like those of the US, UK, and EU have developed policy documents that begin to outline the significant issues regarding regulating AI, but these mostly only acknowledge that there are important social and political issues at hand, and yet do no more than nibble at the margins of the problem. There are alarming instances of AI making racially and gender-based prejudiced decisions on issues as diverse as whether a prisoner gets parole or not, to eligibility for social benefits, credit, employment etc. And of course AI is responsible for increasing displacement of labour, even at the white collar levels. Most of these issues have surfaced in the last two years or so, but the trend is such that massive changes in the next few years are predicted. The issues and questions that the proposal seeks to address therefore are both extremely important as well as urgent. We need a sustained process of dialogue among civil society groups, governments, businesses and the technical community in this regard.
Tag 1: Artificial Intelligence
Tag 2: Social Justice
Tag 3: Regulation
Interventions:
The listed speakers will make some opening remarks and the discussion will then be taken to the round table where everyone will be able to give their views, in two rounds, responding to two sets of questions posed by the moderator. Remote participants will be given an equal chance.
Diversity:
The list of initial speakers have a gender and geopolitical diversity. Since a round table format is being employed, we expect to hear a great diversity of views and perspectives.
Onsite Moderator: Parminder Jeet Singh
Online Moderator: Norbert Bollow
Rapporteur: Nandini Chami
Online Participation:
Online participation will be provided and facilitated, and remote participants given an equal chance to intervene as the physically present ones.
Discussion facilitation:
As mentioned, the subject will first be introduced very briefly by three speakers, and then the moderator will list two set of questions for two rounds of open participation by round table participants.
Conducted a Workshop in IGF before?: Yes
Link to Report: http://www.intgovforum.org/cms/wks2014/index.php/proposal/view_public/198
Additional Speakers:
Agenda:
The onsite moderator will open the session with a statement on the intention of workshop organisers. He will also provide a state of play on what is already being done in the area to ensure that artificial intelligence works for and not against equity and social justice, as is often the fear from many accounts of AI's impact.
Time assigned for the introduction is 5 minutes.
Following the moderator, the 5 subject-matter experts will be given 5 minutes each to give their views and perspectives on the issue described above.
Time assigned for the expert statements is 25 minutes
Following the introduction of the subject, and expert statements, interested persons from the attendees of the workshop will be given the opportunity to make short and focused interventions on the precise statement of the problem. They can also make suggestions/ prescribe directions to be taken to ensure that AI actually works for equity and social justice These proceedings will follow a round-table format and the moderator shall make sure that that interventions do not stray from these two avenues detailed above.
Time assigned for interventions is 50 minutes.
The session will conclude with each panelist detailing key take-aways from the session and indicating how it will influence their work going forward. Each speaker will be given 2 minutes.
Time assigned for the conclusion is 10 minutes
...
